AI Integration: RAG with Gemini API

This showcase adds retrieval-augmented generation with Gemini to my Next.js app. A secure API route calls Gemini with the File Search tool pointed at a managed store, so answers are scoped in my documents or AI training data (e.g., questions about Jakkrit Turner). The UI provides prompt validation, loading feedback, and Markdown rendering for readable responses. Behind the scenes, it involves maintenance scripts to create stores, upload or delete documents, list stores and docs, and retrieve store IDs demonstrating practical, end-to-end RAG setup you can repurpose for support, business profiles, product search and compare, legal/contract assist, and more.

Demo:

Please ask Gemini anything about Jakkrit Turner. For exampe, academic achievements, professional background, and interests.